Oecologia (Berlin) (1985) 65:363-369 Oecologia 9 Springer-Verlag I985 Long-term inducible resistance in birch foliage: triggering cues and efficacy on a defoliator Erkki Haukioja, Janne Suomela, and Seppo Neuvonen Laboratory of Ecological Zoology, Department of Biology, University of Turku, SF-20500 Turku 50, Finland Abstract. Heavy damage of the mountain birch foliage, as well as application of small amounts of insect frass to the soil beneath the trees, reduced growth of Epirrita autumnata larvae reared in these trees in the following year. Foliage damage in the previous year decreased larval survival, too. Both foliage damage and insect frass in the soil decreased a fecundity index which combined the effects of size and survival. Because application of small amounts of fertilizers had an effect indistinguishable from that of insect frass, the effect of the frass may base on responses of trees to an increase in soil nutrient concentration in mid-summer. In previously untreated control trees, all performance indi- ces (growth, survival, and egg production) of Epirrita corre- lated positively with the distance of the birch from the clo- sest birch defoliated in the previous year, indicating "com- munication" between adjacent trees. Epirrita egg produc- tion in trees that had been both defoliated and treated with frass in the previous summer was at least 70% lower than in previously unmanipulated control trees. 1. Introduction Resistance of plants to herbivores is either constitutive or inducible. Inducible forms of resistance, released by herbi- vore action, are interesting because the level of resistance depends on the density of herbivore populations. Therefore, depending on efficacy, and triggering and relaxation times, inducible resistance may tend towards either stabilizing (if triggered and relaxed rapidly) or destabilizing (if either trig- gering and/or relaxation times are long relative to genera- tion time of the herbivore) herbivore populations. High resistance of several species of forest trees has been shown in the growth season(s) following natural or artificial defoliations (Benz 1974; Haukioja and Niemel/i 1977; Werner 1979). Such a long-term inducible resistance can be demonstrated by bioassays showing retarded growth and/or reduced survival of the herbivores. These effects may last for several years (Benz 1974; Haukioja 1982; Tuomi et al. 1984), and they offer a potential causal mechanism for long time lags between population crash and beginning of a new rise in the density of the herbivore. Almost all studies concerning the long-term inducible resistance share the implicit assumption that artificial leaf removal is equally effective with a natural defoliation in Offprint requests to: E. Haukioja triggering the response in the tree. However, Haukioja and Neuvonen (1984) demonstrated that under certain condi- tions artificial and natural defoliation may have different effects. This is not very astonishing because, in a natural defoliation, loss of foliar biomass is accompanied by other cues which might be received by trees and which might affect the quality and/or efficacy of the response by the trees. Such a possibility is especially noteworthy if the in- ducible response of the plant is defensive in evolutionary sense. Consequently, various potential triggering cues and different combinations of them are worth studying when the effects of the defenee on herbivore performance are tried to reveal. We have four main questions in the present paper. First, we report results of experiments which were designed to reveal (1) whether and to what extent manual tearing of foliage triggers long-term defensive responses in the moun- tain birch (Betula pubeseens spp tortuosa) to the autumnal moth (Epirrita autumnata). Use of a manual defoliation technique has a clear advantage: contrary to cases where natural defoliation has occurred, herbivore-spread vectors of diseases cannot be the causal factor causing poor success of animals on induced diets. However, to mimic more close- ly cues which a birch encounters in natural defoliations we also performed experiments where we studied (2) wheth- er application of insect frass in the soil beneath the trees affects performance of larvae on the foliage. Previous defo- liation of host trees was earlier found to lead to retarded growth and lower survival in Epirrita larvae (Haukioja 1980). A pilot experiment had indicated that insect frass in the soil might lead to a similar consequence (Haukioja 1980). Since the pilot experiment we repeated the frass ap- plication experiment twice to find out relative efficacy of defoliation and frass treatments. We also studied (3) wheth- er birches might "communicate" (Baldwin and Schultz 1983; Rhoades 1983), i.e. whether untreated trees might be affected by foliage damage in nearby birches. Finally, we tried to find out (4) how large potential effects the long- term resistance of host trees may have an Epirrita perfor- mance. 2. Materials and methods Epirrita is a univoltine geometrid moth. It overwinters as an egg, larvae hatch in early spring at the time of bud burst in birch, and pupation takes place in (June-) July. Adults fly in August-September. Epirrita larvae are poly-